analysis of human resources for health strategies and policies

Cailhol et al. Globalization and Health 2013, 9:52http://www.globalizationandhealth.com/content/9/1/52

RESEARCH Open Access

Analysis of human resources for health strategiesand policies in 5 countries in Sub-Saharan Africa,in response to GFATM and PEPFAR-fundedHIV-activitiesJohann Cailhol1*, Isabel Craveiro2, Tavares Madede3, Elsie Makoa4, Thubelihle Mathole1, Ann Neo Parsons1,Luc Van Leemput5, Regien Biesma6, Ruairi Brugha6, Baltazar Chilundo3, Uta Lehmann1, Gilles Dussault2,Wim Van Damme5 and David Sanders1

Abstract

Background: Global Health Initiatives (GHIs), aiming at reducing the impact of specific diseases such as HumanImmunodeficiency Virus (HIV), have flourished since 2000. Amongst these, PEPFAR and GFATM have provided asubstantial amount of funding to countries affected by HIV, predominantly for delivery of antiretroviral therapy(ARV) and prevention strategies. Since the need for additional human resources for health (HRH) was not initiallyconsidered by GHIs, countries, to allow ARV scale-up, implemented short-term HRH strategies, adapted toGHI-funding conditionality. Such strategies differed from one country to another and slowly evolved to long-termHRH policies. The processes and content of HRH policy shifts in 5 countries in Sub-Saharan Africa were examined.

Methods: A multi-country study was conducted from 2007 to 2011 in 5 countries (Angola, Burundi, Lesotho,Mozambique and South Africa), to assess the impact of GHIs on the health system, using a mixed methods design.This paper focuses on the impact of GFATM and PEPFAR on HRH policies. Qualitative data consisted ofsemi-structured interviews undertaken at national and sub-national levels and analysis of secondary data fromnational reports. Data were analysed in order to extract countries’ responses to HRH challenges posed byimplementation of HIV-related activities. Common themes across the 5 countries were selected and compared inlight of each country context.

Results: In all countries successful ARV roll-out was observed, despite HRH shortages. This was a result of mostlyshort-term emergency response by GHI-funded Non-Governmental Organizations (NGOs) and to a lesser extent bygovernments, consisting of using and increasing available HRH for HIV tasks. As challenges and limits of short-termHRH strategies were revealed and HIV became a chronic disease, the 5 countries slowly implemented mid tolong-term HRH strategies, such as formalisation of pilot initiatives, increase in HRH production and mitigation ofinternal migration of HRH, sometimes in collaboration with GHIs.

Conclusion: Sustainable HRH strengthening is a complex process, depending mostly on HRH production andretention factors, these factors being country-specific. GHIs could assist in these strategies, provided that they areflexible enough to incorporate country-specific needs in terms of funding, that they coordinate at global-level andminimise conditionality for countries.

Keywords: Human resources for health, Sub-Saharan Africa, HRH policies, Global health initiatives, HIV, GFATM,PEPFAR, Health system strengthening

* Correspondence: [email protected] of Public Health, Faculty of Community Health Sciences, Universityof the Western Cape, Cape Town, South AfricaFull list of author information is available at the end of the article

© 2013 Cailhol et al.; licensee BioMed CentralCommons Attribution License (http://creativecreproduction in any medium, provided the or

Ltd. This is an open access article distributed under the terms of the Creativeommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andiginal work is properly cited.

mailto:[email protected]

http://creativecommons.org/licenses/by/2.0

Cailhol et al. Globalization and Health 2013, 9:52 of 14http://www.globalizationandhealth.com/content/9/1/52

BackgroundIn the early 2000s the international community respondedto the Human Immunodeficiency Virus (HIV) pandemic,in addition to re-emerging Tuberculosis (TB) and on-goingmalaria epidemics, by creating targeted disease-specificfunding mechanisms, termed Global Health Initiatives(GHIs). Among those funding for HIV control were the U.S.President’s Emergency Plan For AIDS Relief (PEPFAR),the Global Fund for AIDS, Tuberculosis and Malaria(GFATM) and the Multicountry AIDS Program of theWorld Bank (MAP) [1]. GHIs were defined as “a blue-print for financing, resourcing, coordinating, and/or imple-menting disease control across at least several countries inmore than one region in the world” [2], pp. 74. GFATMand PEPFAR allocated most funds to antiretroviral (ARV)provision [3,4], which was the most pressing need atcountry-level at that time, without considering humancapacity to deliver these ARVs.Rapid expansion of HIV programmes in Sub-Saharan

Africa (SSA) suddenly revealed the demand for human re-sources for health (HRH) in a continent already depletedof HRH [5,6]. HRH demand was not only quantitative, interms of the scale-up of labour-intensive service-delivery tohundreds of thousands of newly diagnosed patients, butalso qualitative, in terms of new skills needed to managepatients with HIV [7]. SSA was faced with a double chal-lenge in relation to HRH, a structural deficit and an ac-quired shortage. The World Health Organization (WHO)had estimated that 2.3 qualified health workers per 1,000population – 0.55 physicians and 1.88 nurses/midwives -were the minimum, needed to ensure the delivery of 80%of basic services [5]. This minimum threshold, which variesin relation to the epidemiological profile and the productiv-ity of staff, was far from being achieved in most the coun-tries in SSA at the time GHIs launched their activities.The distorting effects of HIV funding on countries’

health systems, including effects on HRH, were being re-ported from soon after the launch of GHIs: PEPFAR, bychannelling their funding through international implemen-ters and GFATM, by implementing their activities throughnational non-governmental organizations (NGOs), wereresponsible for “poaching” staff from an already depletedpublic sector; in many sites, specific units for HIV patientswere created in public facilities around a doctor-centredmodel of ARV provision, fragmenting the health system;and the skills needed for HIV management were initiallydeveloped in standalone programmes and in an uncoordin-ated fashion, mostly through in-service trainings [8-13].Such country-level findings, which were often based on

key informant interviews at the national level, brought theHRH issue to the attention of global stakeholders and be-came incorporated into the “health systems strengthening”global agenda. This was evidenced by a number of glo-bal initiatives on HRH which multiplied during the last

decade: Joint Learning Initiative [14], World health Re-port [5], Global Health Workforce Alliance, JapaneseInternational Cooperation Agency’s commitment totrain additional health workers, renewal of interest inCommunity-Health Workers and Mid-Level Workers,WHO global plan to “treat, train and retain” HRH [15].At country-level, a number of innovative approaches

and new policies, different in nature according to thecontext, were also developed in response to the aggra-vated HRH crisis.This paper provides an analysis of both short-term

and long-term HRH responses to implementation ofPEPFAR and GFATM funded activities, in 5 countries inSSA. It draws on health policy and system research, anapproach born from the necessity to analyse systematic-ally globalized and complex health systems.

MethodsA five year multi-country study (2007–2011), funded bythe European Commission was conducted in 5 countriesin SSA (Angola, Burundi, Lesotho, Mozambique andSouth Africa), to describe and analyse the effects ofGHIs on country health systems. GHIs common to all 5countries were PEPFAR (except Burundi), MAP (exceptSouth Africa) and GFATM. MAP was excluded from thestudy since not all countries could collect data related toit. The research focused mainly on countries’ responsesto the increased demand on HRH due to the challengesposed by GHI-funded HIV activities.The research design was a mixed methods approach,

based on a conceptual framework developed to analyse theinfluence of external aid on countries’ health policy andsystems. Tools for quantitative and qualitative data collec-tion were jointly developed by the multi-country team,covering the 6 building blocks of the health system [13].Details on each country’s method for data collection

and analysis are presented in Table 1. Ethical approvaland permission to conduct research were obtained fromrelevant authorities in all countries prior to data collec-tion. Sampling, from provinces down to facilities, wasdone by each country team, and finalised following con-sultation with relevant authorities.Qualitative data consisted of document review and

semi-structured interviews, conducted at different levelsof the health system. A first group of respondents was se-lected according to their relevance to the topic researchedand thereafter more were recruited using the snowballtechnique. A total of 145 interviews at national-level and419 at sub-national level were conducted in the 5 coun-tries. National-level interviews included key-informantinterviews at national and sub-national levels and of devel-opment partners (international and local) and governmentofficials.

Table 1 Methods used and details of data collection and analysis in each country

Angola Burundi Lesotho Mozambique SA

Ethicsapproval

National and provincialhealth authorities

National ethicscommittee

Ethics committee of theMinistry of Health andSocial Welfare

National bioethicscommittee

National, provincial andmunicipal governmentresearch committees,UWC ethics committee

Period ofdatacollection

2009 April-June 2009 February- June 2008 July- 2009 February 2007 March- September;2008

2008 September −2010October

2010 May 2011 March- June 2010 February-May

2011 June-September

Languageof datacollection

Portuguese French / Kirundi English Portuguese and English English, Xhosa, Afrikaans,Zulu

Numberand type ofinterviews

National level: Ministry ofHealth (minister, advisorsfrom the Ministry ofHealth / PAV-MINSA“immunization program”),and offices of selectedImplementing Partners(UNICEF, ONUSIDA, EU,WHO) and NGOs 11 in2009 1 in 2010

National level: MoHofficials (seniormanagers, HRH/planning/ programsmanagers), National AIDSCouncil secretariat, NGOsrepresentatives, GHIsrepresentatives 27 in2009 26 in 2011

National level : in-depthkey informant interviewswith 22 representativesof the government,bilateral and multilateraldevelopment agenciesand other stakeholders atthe national level

National level: 21 in 2008;Ministry of Health (MoH)officials, offices ofselected ImplementingPartners (WHO, UNAIDS,UNICEF, Irish Aid, PMI/CDC, World Bank, USAID,CDC, DFID) and NGOs(MONASO-network ofnational NGOs workingon AIDS, MalariaConsortium, HealthAlliance International

National level: MoHofficials, SANAC,international NGOcoordinators,international healthagencies coordinators 19in 2008–2009 18 in 2010

Sub-national level:provincial governmentofficers, NGOs, districtmanagers, facilitymanagers 30 in 2011

Sub-national level:provincial governmentofficers, provincial AIDScommittee, NGOs,district managers, facilitymanagers andemployees 35 in 2009 45in 2011

Sub-national level:around 60 with provincialand district healthdirectorates, HRmanagers, NGOmanagers, individuals incharge of health facilitiesand services responsiblein 2010

Sub-national levels:Provincial and municipalgovernments, sub-provincial managementlevels (general manager,HR manager, financemanager) N = 105 from2009 to 2010

Facility and NGO level:NGO representative,health workers, facilitymanager N = 144

Type ofdocumentsanalyzed

National health policy,national programs ofMalaria, Tuberculosis andMaternal Health andnational and internationalreports; Publishedliterature andunpublished documentsprovided by keyinformants of MINSA,national and internationalNGOs, Provincial HealthDepartment andMunicipal Hospitals.

Policy and planningdocuments fromprograms and HRH unit,national health plan,proposals for GHIs

Review of Lesotho’sRound 5 Grant ScoreCards Grant Performancereports and policy andplanning documentsfrom programs and HRHunit, national plan : 10-15

Policy and planningdocuments from Nationalprograms and HRNational Directorate,national health plan, greydocuments

Policy and planningdocuments from nationaldepartment of health,grey documents,proposals for GHIs, draftpolicies

Quantitativesamplingmethod(sub-nationallevel

NA 3 provinces (2 rural andone urban, in each ofwhich 3 or 4 facilities(NGO, private and public)were selected (14 in2009 and 12 in 2011)

NA NA 3 provinces, minimum of2 districts in each,minimum of two facilitiesin each district at lowestlevel providing ARTinitiation. Rural/urbansampling where possibleat each level

Type ofquantitativedata

HRH national report andnational HIV programreport for ART patientsfigures

Survey at facility level(N = 105 in 2009, N = 78in 2011) to health workers(salary level, incentives,trainings, supervision)

NA Surveys at facility level forHRH, infrastructuresmapping, healthinformation system, andpharmacy information;

Health system trustdatabase for HRH data;

National ART report forART patients number


Table 1 Methods used and details of data collection and analysis in each country (Continued)

surveys for NGOmapping and districthealth services network

Surveys at facility levelfor HRH number andtrainings, surveys forNGO mapping anddistrict health servicesnetwork

HRH national review andnational HIV programreport for ART patientsfigures

Analysismethod(softwareused)

Qualitative data: Thematicanalysis - analyzedmanually

Qualitative data:Framework analysis usingAtlas.ti for

Qualitative data:Framework analysis usingAtlas.ti

Qualitative data:Thematic analysis usingNvivo and contentanalysis

Qualitative data:Thematic analysis bothmanually and usingAtlas.ti

Quantitative data:analyzed using Stata(version 8)


Interviews were recorded whenever possible, tran-scribed, translated when needed and coded. Writtenconsent form was obtained prior to each interview. Datain each country were analysed by country teams, consti-tuted of at least 2 researchers, using thematic analysis,done either manually or with software. Annual work-shops were conducted to discuss data collection chal-lenges and preliminary findings, to allow triangulation ofdata and to identify relevant common themes.Quantitative data consisted of surveys at facility level

and secondary data from government information sys-tems. Two categories of HRH in the public sectorwere considered for quantitative data: medical doctorsand professional nurses (basic/intermediate/specializedlevels). Other categories, such as pharmacists and com-munity health workers were not homogeneous enoughacross countries to allow comparisons, or data weremissing which precluded cross-country comparisons.Specialist doctors were merged with general medicaldoctors. Nurse and midwife figures were merged whenpossible to allow comparisons with the health workerthreshold above cited [5].We provided quantitative data on evolution of num-

bers of HRH in the public sector (2004–2010) and onARV roll-out (2004–2009).This paper first presents and compares background

data on each country, followed by the short-term HRHstrategies adopted and challenges they posed. The paperthen examines countries’ transitions from a short-termstrategy to long-term HRH policy, as HIV became achronic disease. Short term strategies were defined asstrategies implemented by local stakeholders without aformal regulatory framework, mostly via internal ar-rangement. Long-term policies were defined as thosebacked-up by an official authority and a regulatoryframework, allowing standardization and sustainability.We finally discuss similarities and differences in strat-egies and policies across countries, in light of their con-text and more generally, the HRH strengthening needsand the role of GHIs.

ResultsCountry contextsRelevant features of the 5 countries studied are presentedin Table 2. All 5 countries now have democratically electedgovernments, though most are relatively new multi-partydemocracies, including South Africa and Mozambique(1994) and Burundi (2005). Burundi gained independencein the 1960s, whereas Angola and Mozambique did so in1975. These 3 countries have experienced civil wars,Mozambique just after its independence and more re-cently Angola and Burundi. In terms of demography,Burundi is an outlier, with 318 inhabitants per km2,whereas other countries’ densities range from 15 to 71 perkm2. South Africa is an upper-middle income country,Lesotho and Angola are lower-middle income countriesand Mozambique and Burundi are low-income countries(World Bank classification 2009). The position of the 5countries within this classification was correlated with theamount of Official Aid for Development as % of Gross Na-tional Income: Burundi and Mozambique are the 2 coun-tries most dependent on external aid.In terms of burden of HIV, Lesotho and South Africa

carry the heaviest ones, with, in 2009, an antenatal HIVprevalence of 23.6% and 17.8% respectively. South Africa’slarger population meant it had more than 5.5 millionHIV-infected people in 2009 and almost half a millioncases of TB diagnosed in 2010. Angola and Burundi havethe lowest incidence of TB (304 and 129 cases per 100,000inhabitants respectively in 2010) and HIV prevalence (3.3and 2% respectively in 2009). Angola, Burundi andMozambique have the worst indicators for maternal andchild health.With regard to health financing, all governments spent

in 2009 from 8 to 13% of their total expenditure on thehealth sector; this percentage included funding from ex-ternal sources. The latter represents as much as 72% forMozambique, 45% in Burundi, and 30% in Lesotho, butonly 3 and 2% for Angola and South Africa respectively.If external sources of funding are excluded, governmentexpenditure on health would drop significantly in the 4

Table 2 Summary of socio-economic and health related indicators for the 6 countries included in the analysis

Angola Burundi Lesotho Mozambique SouthAfrica

Significant historic features Independence 1975 Independence1962

Independence1966 Independence:1975

Apartheid1948-1994

Quarter century ofcivil war: 1975 - 2002

Cyclic civilwars since1963

Several military coups with latesthandover to democratic government in1995

Civil war: 1976-1992

Latest: 1993-2006

Population density/km2 15 318 71 29 41

Net ODA as% of GNI 0.3 42.3 5.4 20.8 0.4

GDP in current USD per capita 4069 163 800 428 5733

Public health expenditure, as % oftotal government expenditure

8 12 8 13 9

External resources for health, % oftotal expenditure for health

3 45 30 72 2

OOP expenditure on health, % oftotal expenditure for health

11 36 22 12 18

GINI coefficient (latest available) 58.6 (2000) 33.3 (2006) 52.5 (2003) 47.1 (2003) 67.4 (2006)

Human Development Indexranking 2011 (out of 187countries)

148 185 160 184 123

HIV prevalence 15–49 years old, %(2009)

2 3.3 23.6 11.5 17.8

Number of persons affected byHIV, all ages, 2009

200,000 180,000 290,000 1,400,000 5,600,000

TB incidence, per 100,000inhabitants, 2010

304 129 633 544 981

Number of TB cases detected,2010

58,000 11,000 14,000 130,000 490,000

Malaria mortality rate per 100,000inhabitants, 2008

89 39 0.1 171 0.2

Under-five mortality rate, per 1,000live births, 2010

161 142 85 135 57

Maternal mortality ratio per100,000 live births, 2008

610 970 530 550 410

GFATM-HIV, cumulativedisbursement, as of 2011, millionsUSD

62.2 69.5 91.8 168.7 247.6

GFATM-malaria, disbursed, as of2011, millions USD

62.0 55.2 0 61.6 0

GFATM-TB, disbursed, as of 2011,millions USD

10.3 9.8 10.7 12.6 0

PEPFAR, disbursed, as of 2009,millions USD ( + committed 2010)

47.7 0 (Not eligible) 96.2 1096.7 3113.4

MAP1 World Bank, committed,millions USD

21 51 (withMAP2)

5 55 0

Sources:Socio-economic and finance indicators; World Bank database, 2009.Health indicators: WHO global health indicators database.Disbursements indicators: GFATM website, PEPFAR country-websites, World Bank MAP-country websites.Human development index: UNDP.



first countries, and would then represent less than 7% oftotal government expenditure.The health system organization is typically pyramidal in

all countries, with a centrally managed district-based sys-tem. In Burundi, decision-making remains very centralizedat national level, especially budget allocation, since the dis-trict health system is very recent. In South Africa, the na-tional level plays a normative role, by drafting policies andallocating funding, while provinces have considerable au-tonomy and exhibit variation. In Angola, the national levelhas a normative role and since 2007, municipal health au-thorities are in charge of district health management (in-cluding budget allocation) and planning. In Mozambiqueand in Lesotho, the decentralization process has beengradual. In Mozambique, provinces gained autonomy,while budget allocation is still centralized. All countrieshave HRH development policies and plans, but their levelof ownership and sophistication varies: South Africa’s planis relatively detailed with projections of HRH until 2025[16]; Mozambique’s is clearly articulated with MDGs andaligned with other policies but lacks funding (NationalPlan for HRH Development 2008–2015; National Direct-orate of Human Resources, Ministry of Health, 2008);whereas Burundi launched its first plan only in 2010 [17];Lesotho’s HRH policy is clearly articulated with the Healthand Social Welfare Policy (2004) which ensures appropri-ate supply, and properly trained personnel to meet theneeds of the country and also ensures career development.In terms of HRH quantity, all countries except South

Africa have experienced a general HRH deficit [18]. How-ever, these figures are aggregates, neglecting the break-down between public and private for profit (PFP) staff. InSouth Africa, the PFP sector is favoured by physicians,with 59% of them working in this sector in 2010, whileserving only 15% of the population [19]. In the other 4countries, the PFP sector is not so prominent, though itsassessment is difficult, since many HRH working in the

Figure 1 Evolution of selected HRH indicators in the public sector, in th

public sector are in dual practice, undertaking a parallelPFP activity. If only the public sector is considered and thethreshold of 2.3 health workers (0.55 doctors and 1.88nurses and midwives per 1,000 population) applied, all 5countries have unmet needs, both in 2004 and 2010(Figure 1). The extent of unmet needs in the public sectorvaries, with the lowest density of physicians and pro-fessional nurses in Burundi and Mozambique, whereasAngola has a shortage of physicians and has nurses in ex-cess (though most of them have only basic level education)when compared to the WHO threshold.Since at least 2004, all 5 countries have been receiving

funding from the 3 GHIs (GFATM, PEPFAR -except forBurundi- and the World Bank MAP -except for SouthAfrica-).The amounts disbursed by PEPFAR greatly out-weighed those of GFATM and MAP, especially in SouthAfrica and Mozambique (Table 2), much of it for ARVroll-out. Lesotho was not a focus-country for PEPFARand Burundi was not eligible for PEPFAR until 2011. Interms of subsequent ARV roll-out, the increase in thenumber of patients on ARV was dramatic in all 5 coun-tries between 2004 and 2009 (Table 3), despite the sus-tained deficit in HRH (Figure 1). The majority ofpatients were enrolled for ARV in the public sector, ex-cept for Burundi where the majority were initially en-rolled in stand-alone HIV-clinics run by local NGOs.Data from our study showed that this successful ARVroll-out, despite a HRH deficit in these countries, was aresult of adaptive practices, which occurred in a similarway in all countries: initially, a short-term emergency re-sponse mostly by GHI-funded NGOs and to a lesser ex-tent by government; then, challenges related to thisemergency-type response increased and, as HIV becamea chronic disease, government started to formulate andimplement long-term policies, sometimes in collabor-ation with GHIs. This 2-step country-response is de-tailed below.

e 5 countries, between 2004 and 2010.

Table 3 Numbers of patients on ARV in 2004 and 2009 in the 5 countries

Year Angola Burundi Lesotho Mozambique South Africa

2004 (unless otherwise specified) 5,357 1,200 54,237 8,010 32,895 (2005)

2009 (unless otherwise specified) 20,640 17,500 (2010) 92,773 (2008) 134,147 781,465

Sources:Angola, Burundi, Lesotho, Mozambique: national HIV program reports.SA: ART factsheet from national department of health.


Short term strategies: facing the HIV emergencyUsing the pool of available HRH for HIV tasksInformal or formal task-shifting at facility-level In allcountries except Angola, task-shifting (to less skilledHRH or to lay health workers) occurred at facility-levelfor HIV clinical management. In South Africa, it wasmainly an initiative of international NGOs, first infor-mally by Médecins Sans Frontières (MSF), during theperiod of HIV denialism in the government [20],followed by GHI-funded NGOs. In Burundi, informaltask-shifting was in place in rural areas since before thecivil war, because of pre-existing doctors’ shortage. How-ever, task-shifting for HIV was not considered initially.More recently, informal initiatives with follow-up ofARV patients by nurses were conducted as pilots, mostlyby GFATM-supported NGOs. In Mozambique, the Min-istry of Health adopted HIV task-shifting policy early on:formal non-physician clinicians pre-dated the GHIs eraand have been further trained for ARV prescription; also,in some rural facilities, nurses were trained to follow-uppre-ARV patients. In Lesotho, task-shifting has alwaysbeen supported by the government. Nurses in the ruralareas have always worked under a written protocol bythe district medical officer. This has been extended toHIV/AIDS; nurses have been trained further to initiateARV treatment in uncomplicated cases.Challenges: In South Africa, facility-level task-shifting

was well accepted locally but remained for a while asscattered pilot projects; and its scale-up was slow, due toa lack of formal process and standardization and an ac-tive opposition from professional councils. In Burundi,its uptake was slow, mainly due to some resistance tochange from professional bodies as well as from patients,since HIV was considered a complex disease. InMozambique and Lesotho, where task-shifting was oper-ating historically since before GHIs, its adaptation to in-clude HIV activities was easier.

Allocation of HIV-tasks to specific health workers Inall countries except Angola, specific units for HIV servicesopened in public facilities. Separation of HIV and non-HIV activities aimed at tighter management of externalfunding and at training a limited number of HRH. Newstaff were recruited or staff formerly present were intern-ally redeployed to work in these units. Sometimes one staffmember was allocated to HIV-specific tasks within the

same unit (e.g. Prevention of Mother-To-Child Transmis-sion -PMTCT- in Maternal and Child Unit).Challenges: In the short-term, this system allowed imple-

mentation of urgent and specific activities but proved un-sustainable in the medium-term. Indeed, the numbers ofspecific HRH needed (clinical as well as administrative)had to be increased exponentially, in parallel to patientnumbers. In Mozambique, this stand-alone approach wasdiscontinued in 2007 and HIV services were progressivelyintegrated into the public health sector. Also, except inSouth Africa, staff within these specific HIV units and insome cases staff allocated to specific HIV tasks withinthe same unit received supplementary stipends out ofGFATM-HIV grant. These practices induced team divisionsand tended to undermine mainstreaming of HIV activities.

Increasing the pool of HRH for HIV tasksUse of retired staff (only in Lesotho) Retired nursesare used as mentors for PMTCT (by internationalNGOs) - administration of that program is being movedto the government with local mentors being paid by thegovernment. This strategy has worked because Lesothowas in dire shortage of nurses and utilized available andexperienced staff. The biggest challenge was the remu-neration while nurses were transferred back to govern-ment, since international NGOs paid more than thegovernment.

Stand-alone GHI funded NGOs for HIV-activities Inthe case of PEPFAR and GFATM, NGOs deliveringHIV-care were either direct or indirect recipients offunding, as disbursement through civil society was aconditionality to get their funding [21,22]. In all coun-tries, local NGOs were mainly funded by GFATMwhereas PEPFAR implementers, who subcontracted tolocal NGOs, were often U.S.-based. In Burundi andMozambique, GHI implementing agencies, separatefrom government, were created. In Burundi, these agen-cies also funded local NGOs and some of these in turncreated HIV-clinics, run independently from govern-mental institutions. In all countries, GHI-funded NGOsoffered higher salaries than the public sector, allowingthem to attract more qualified staff. In Burundi, onaverage, GFATM-funded NGO staff was paid 2 to 3times more than in the public sector at the time of thesurvey at facility-level (quantitative data, 2009). In


Mozambique, Angola and Lesotho, key informants allreported that PEPFAR implementers provided muchhigher salaries than GFATM-funded NGOs.Challenges: All countries faced an internal brain-drain

of experienced staff from the public sector to GHI-funded NGOs, due to differences in salary level. InBurundi and Lesotho, migration was also observed fromthe public sector to the National AIDS Council (NAC)secretariat, where management staff received highersalaries than in the public sector.

HIV staff seconded to facilities - (in South Africa only)A number of PEPFAR implementers were allowed bygovernment to second staff to public sector HIV clinics,to help them with accreditation and to then hand overthe clinic to the facility in which they were located. Inone province, GFATM-funded HRH, dedicated to HIV,were placed at facility level and directly managed by theprovincial department of health.Challenges: Clinical staff, paid by PEPFAR implemen-

ters and seconded to facilities, were officially paid at thesame rate as public sector staff, as a result of internal ar-rangement, but had higher net incomes as governmentbenefits were not automatically deducted, incurringsome resentment from the rest of the staff. The verticalmanagement system of GFATM created confusion andundermined integration. Concerns around whether theseseconded staff would be absorbed by the public sectorwere being raised, as GHI funding began declining.

Informal task-shifting at community-level In all coun-tries, Community Health Workers (CHWs) existed beforethe GHI era. In South Africa, CHWs were further re-cruited and trained, mostly by local NGOs subsidized byGFATM grants, to provide prevention and promotionactivities on HIV at community level. In Burundi,Mozambique and Lesotho, new cadres of health workerscalled “health mediators”, “lay counsellors” or “expert-patients” were introduced through funding from GHIs,dedicated exclusively to HIV-related activities [23]. Theirtraining has been provided by GHI-funded NGOs(Lesotho). After their training, they receive short-termcontracts with NGOs or hospitals, working under thesupervision of NGOs and receiving GHI-funded stipends.Challenges: In contrast to implementation of task-

shifting at facility-level, community-level task-shiftingworked well, since it preceded the emergence of theGHIs and did not interfere with existing power and hier-archies amongst professional HRH. However, inBurundi, the curricula of the new cadres of HRHremained ‘unofficial’ rendering their situation precarious.“Health mediators” were left unpaid for months, due toconfusion during the transition between 2 GFATMrounds [24]. The lack of harmonization of trainings

amongst different training providers was also a concernin all countries.

HIV as a chronic disease: towards long-term policiesFormalisation of pilot initiativesRegulatory framework for task-shifting at facility-level Informal task-shifting was followed by attempts tolegitimise expanded professional responsibilities and tostandardize practices. In South Africa, the governmentauthorized formal nurse-initiated antiretroviral therapy[25], after the STRETCH (Streamlining Tasks and Rolesto Expand Treatment and Care for HIV) trial demon-strated the safety of such a policy [26]. In Burundi in2008, a ministerial circular authorized nurses to pre-scribe ARV, though under the supervision of a medicaldoctor. In Mozambique and Lesotho, where task-shiftingpre-dated GHIs, scopes of practice of non-physician cli-nicians and nurses were officially expanded, with HIVtraining provided by PEPFAR and international imple-menting partners (e.g. International Training & Educa-tion Centre for Health in Mozambique).

Official creation of new cadre of health workers Twonew cadres of health professionals were officially createdin South Africa in 2008 with their job description re-leased by national level through a circular in 2010[27,28]. The curriculum of one of them, called clinicalassociates, includes also HIV management and is par-tially supported by PEPFAR (funding of HIV/AIDS twin-ning centre partnerships with US institutions). Amiddle-level pharmacy cadre, called pharmacists' assis-tants, has also been created, with some initial resistancefrom the pharmaceutical bodies. CHWs are recognizedin South Africa as having significantly supported ARVroll-out. With the latest, far-reaching reforms in SouthAfrica, known as 'Primary Health Care (PHC) re-engineering’, CHWs are now being integrated into so-called “ward-based PHC outreach teams”, working underthe supervision of nurses and as part of the formalhealth system [29,30].

Increasing the overall quantity of HRHIncrease in production In reaction to overall HRHshortages, all governments made plans to increase localproduction, sometimes in collaboration with GHIs. InSouth Africa, nurses’ colleges which had been closeddown were due to reopen, as part of the new HRH plan,and production by medical schools was due to increase.In Angola, 6 public medical schools were established in2009 in the provinces, in addition to a pre-existing pri-vate school. In Burundi, production of doctors andnurses has increased due to an increase in student intakebut also with the creation of private schools, though


without a clear linkage between Ministries of Health andEducation.In general, the capacity of pre-service training institu-

tions was already overstretched before the HIV crisis,raising concerns about the quality of graduates. Theseproduction increases were sometimes supported by bilat-eral cooperation: in Lesotho, Irish Aid contributed totraining 150 additional nurses in year 2007 while somenurses were recruited from Kenya to work in the remoteareas of the country; South Africa has a bilateral agree-ment with Cuba, to train doctors. PEPFAR supportsHRH production since 2011 (in South Africa andMozambique through Medical Education PartnershipInitiative-MEPI- and in Lesotho through Nursing Educa-tion Partnership Initiative-NEPI-) [31,32].

Mitigation of internal migration induced by inequalities insalaries

Anti-poaching agreements South Africa was the onlycountry where evidence of informal policy adopted by anumber of NGOs to stop the poaching of HRH from thepublic sector was reported (in one province only). In thiscase, NGOs signed an agreement with provincial author-ities to not recruit public sector HRH from the sameprovince. However, it did not prevent poaching fromother provinces.

Alignment of NGO salaries to those of public sectorAs a result of significant salary inequity between GHI-funded NGOs and the public sector and the subsequentbrain-drain, salary harmonisation was attempted in mostcountries, driven by government or NGOs. In Angola, aformal and mandatory alignment of NGO salaries withthose of the public sector was introduced via a decree.As a result, migration of HRH from the public sector toNGOs was partially reversed. In South Africa, alignmentoccurred through informal and local agreements, withconsiderable variation between different local govern-ments. However, such salary alignment was only for clin-ical or monitoring and evaluation M&E staff (and notmanagement level staff): some NGOs proactively enquiredabout salary levels in the public sector, before setting theirown scales upon recruitment. No formal agreement existedat higher level, i.e. provincial or national level. InMozambique also, in 2006, the Ministry of Health imple-mented by decree a new policy on salary harmonizationbetween public health sector and NGO staff seconded topublic health facilities. As a result, personnel from NGOsare no longer seconded to facilities but give external sup-port to the public health sector. This policy does not go be-yond the staff seconded to facilities and as such does notaddress income inequalities between external NGOs andpublic health sector employees. In Lesotho, NGOs and

other implementing partners still have their own salaryscales, independent from the public sector.

Increase in public sector salaries and provision of in-centives In Burundi, the government was unable to en-force alignment of NGOs’ salaries to public sector ones,since this latter were outrageously low. In 2010, eventually,industrial action forced an increase in public HRH salarylevels, which were aligned to those in neighbouring coun-tries (i.e. Rwanda). Government also implemented a reten-tion policy, including financial incentives to public sectorHRH allocated to rural areas. Migration of HRH was thenreversed at the same time as GHIs funding stagnated. InLesotho, although salaries remained low, the governmentinstituted incentives for nurses in the public sector. Thiswas initially meant for nurses who worked with HIV/AIDSpatients through the support of donor funding. The allow-ance was subsequently spread to cover all nurses includingthose working in hospitals. In Mozambique, the govern-ment introduced incentives for senior managers to preventthem from migrating to NGOs in response to evidence ofsuch internal migration [33]. There was a similar policy inBurundi, where the financial top-up came out of GHI orbilateral projects. In South Africa, no evidence of GHI-funded incentives was found.

Redistribution of incentives at facility-level Burundiand Lesotho, where some specific categories of staff werereceiving incentives for TB or HIV activities, piloted anew policy of redistributing those incentives within theentire facility (a decision of NAC in agreement withfacility-managers in Burundi, a decision of governmentin Lesotho). It did not resolve the issue of low salariessince it resulted in a very small amount per person.

Participation of GHIs in performance-based financing(Burundi only) From 2010, GFATM-HIV started to sup-port the national policy on improving HRH workingconditions using a performance-based funding system.GFATM funding served partly to pay for performancerelated to HIV activities. The policy was mainly donor-driven and donors managed to persuade other signifi-cant funders to pool available funding.

DiscussionThis multi-country analysis shows that all 5 countriesfirst adopted short term strategies in reaction to the HIVpandemic and subsequent GHI funding. Those strategiesconsisted mainly of using the pool of available HRH todeliver on the increased volume of HIV tasks, notablyincreased numbers on ARVs. These strategies includedtask-shifting at facility-level, staff attraction and reten-tion using incentives, and creation of stand-alone HIVclinics. To a lesser extent, the pool of HRH was


increased through creation and recruitment of newcadres of health workers for HIV, use of retired staff,and secondment of NGO staff to public sector facilities.On the one hand, these strategies allowed a substantial

increase in the number of patients on ARV within a fewyears and fulfilled the primary objective of HIV pro-grams and funding institutions. However, on the otherhand, these strategies generated tensions: between HIV-dedicated and general HRH staff, by creating parallelchannels of management and inequities in incomes.While HIV-dedicated HRH positions were better paid,these positions were not sustainable and tasks were notalways clearly defined.Eventually, countries favoured certain short-term HRH

strategies over others, with support from GHIs andother donors.

Short-term strategies: implemented according to GHIs’common funding conditionality and countries’ specificcontextsIn all countries, funding from GHIs provided a substan-tial amount of money, usually with a clear conditionalityrelated to quantitative targets achievement, to kick startthe disbursement of the next financial tranche [21]. Thismechanism of funding was new, placing significantpressure on country systems and accountability. Thispressure contributed to the creation of parallel HIVunits to enable M&E reporting to GHIs, and to allowtargets to be reached quickly [9]. The necessity toquickly reach such targets, led to recruitment and train-ing of HRH, who were retained using incentives. Re-cruitment however occurred only in a limited number,due to the restricted funding amount. The creation ofvertical programs was seen to be justified by the sense ofemergency, generated by a series of high level global pol-icy meetings around 2001–2002 in response to thegrowing HIV epidemic in SSA. It probably also served todemonstrate the effectiveness of certain types of fundingover others, thus fuelling pre-existing competitionbetween agencies at global level [34]. In addition to theworkload added by HIV care, the combination of pre-ventative and curative care that GHIs funding was meantto deliver was new to public sector facilities, and GHIsproposals were an open call to NGOs and civil societyparticipation [21]. This partly explains the extensive useof NGOs (local and international) as short-term strat-egies to overcome the HRH crisis, either by secondingtheir staff to facilities, or by creating stand-alone NGOsfor HIV care. In addition to these explanations above,which influenced certain types of short-term HRH strat-egies over more long-term ones, each country reflectedits specificity, according to its historical and politicalbackground.

In South Africa, the health system in the pre-GHI erawas already sufficiently organized relatively to the 4other countries. On the one hand, this allowed second-ment of NGO staff to public sector facilities, accordingto a pre-defined agreement. On the other hand, the veryhierarchical and managerial functioning of the South Af-rican health system did not immediately allow a cleartask-shifting model to be implemented at facility-level.NGOs were the most innovative in this domain, testingseveral models of care in different facilities.In Burundi, Lesotho and Mozambique, a parallel

implementing agency, separate from the Ministry ofHealth, was created to manage funding from GHIs. Fur-thermore in Burundi, a separate Ministry of HIV/AIDSwas created. Such separation between health and so-called HIV sectors, while facilitating the daily manage-ment in the short-term, created inequities in treatmentbetween HIV-dedicated and general staff. Also in these 3countries, the use of CHWs was well accepted, sincerural areas have relied on them for decades, especiallyduring times of civil war. In Angola, no task-shifting wasemployed at facility level, nor were HIV units created,possibly related to a relatively low prevalence of HIV inthe country.As challenges and limits of short-term HRH strategies

were revealed, the 5 countries slowly implemented midto long-term HRH strategies, such as formalisation ofpilot initiatives (new cadre of health workers, task-shifting), increase in HRH production, and mitigation ofinternal migration of HRH, by increasing public HRHsalaries or by imposing salary alignment on NGOs.

The move of countries towards Health SystemStrengthening (HSS), including HRH strengtheningContent of HRH strengthening policy differed, accordingto, inter alia, countries’ financial and political resourcesand to their resultant degree of dependence on externaldonors. Thus, South Africa and Angola could more eas-ily create new cadres of health workers since these gov-ernments had sufficient funds to devote to indigenoustraining institutions and to increase production of HRH.Some countries, such as Burundi, did not have the nego-tiation space to impose a code of conduct on NGOs (i.e.to set a salary scale). Such policy space depends signifi-cantly on the extent of reliance on external donors, in-cluding NGOs. Burundi preferred to increase publicsector salaries instead, which was already a pressingissue. South Africa managed to impose an anti-poachingagreement on NGOs and both Angola and South Africamanaged to force NGOs to align their salaries with pub-lic sector salaries. The Mozambique government forcedto align salaries of NGO staff seconded to public sectorfacilities with those in the public sector, but did not haveany control over salaries of NGO staff per se.


Countries also implemented HRH strengthening pol-icies, independently of GHIs: in Angola, South Africa andBurundi, governments started to sign bilateral agreementswith countries which have a surplus of HRH, such as Cuba(for South Africa, though this pre-dated GHIs), China andEgypt (Burundi). In all countries except Angola, govern-ments implemented incentive schemes for rural areas andsometimes retention schemes for undersubscribed special-ties, independently of GHIs. In South Africa, the nationaldepartment of health introduced increased remunerationto some categories of public sector HRH to reduce exter-nal or internal migration (‘occupational specific dispensa-tion’). Medical interns are also required to performcommunity service for 2 years in rural areas. InMozambique, health workers working in remote ruralareas are allocated an “isolation subsidy” which increaseswith remoteness of the placement. In Lesotho, to increaserural coverage, public sector nurses are mandatorily sentto rural areas for a period of 2 years (national policy inplace since before GHIs). Also, some faith-based organiza-tions, mostly located in rural areas, provided lower salariesthan government and hence are being subsidized by gov-ernment. The government, with donor support, recruitednurses from Kenya and Zimbabwe to work in some ofthe health facilities in the remote areas. Since 2006, inBurundi, medical doctors need to spend a 2-yearmandatory period in the public sector, mostly in ruralareas, if they wish to specialise.

What were the drivers for changes in GHI policies?GHIs and governments increased their collaboration andGHIs reshuffled their scope of activities in order to fitthe HRH strategies at country-level, following the Parisdeclaration and acknowledging the need for aid effect-iveness and increased coordination. GHIs started to takeHRH pre-service trainings into consideration, althoughproduction was not initially their concern. PEPFAR forinstance launched the MEPI and NEPI in 2010, two pro-grams aiming at strengthening pre-service training ofHRH [35].Whether these changes were induced by repeated criti-

cisms of GHI’s targets which were predominantlyaligned to short-term objectives, such as short-coursesand in-service trainings, has been a subject of debate[36]. Attribution to one or other cause is not possibleand these changes were probably a result of a mixture ofevents and influences: GHIs were evaluated through apre-set mechanism (i.e. 5-year GFATM evaluations forGFATM) [21], and further analysed via independent orcommissioned research; many of them emphasized thefragmenting effects GHIs were having on the health sys-tem and in particular on HRH [9,11,13,37-39]; externaladvocacy towards more focus on HSS and against “AIDSexceptionalism” also grew, through conferences and

publications; a global HSS debate was complemented bya focus on long-term HRH production [40].As countries’ responses to HRH challenges were re-

ported to GHIs, some of these, such as GFATM and GlobalAlliance for Vaccines and Immunization (GAVI), with aparticipative approach, allowed criticisms to be partly ad-dressed in a timely manner: GFATM in 2005 introduced aHSS section to their funding application and GAVI-HSSwas launched in 2006 in pilot countries [41]. Other GHIs,such as PEPFAR, more rigid in its functioning and admin-istration, also took into account such criticisms, but at aslower pace and in a more contained way [22,42]. The slowshift to more sustainable, indigenous responses towardsHSS and especially HRH strengthening seems therefore aresult of a dialectical relationship between GHI-driven ini-tiatives and country responses, while the shift has also beenimpelled by other objective changes (e.g. availability of gen-eric drugs). Reduction in global funding is also certainlyforcing countries and GHIs to invest more in long-term ac-tion, such as pre-service training, better coordination oftraining, collaboration across sectors through partnershipsor across programs (TB-HIV, HIV-reproductive health)and equity in salaries.

HSS and HRH strengthening: what’s in the name?While GHIs should be commended on their genuine ef-forts to improve countries’ health system, whether theseHSS plans are really strengthening or just supporting thehealth system is a matter of debate [43]. Also, an ideo-logical struggle seems to have arisen around whether GHIshave de facto strengthened health systems or had the op-posite effect [44-48]. Answers to these questions will influ-ence future HSS directions for GHIs and countries andcaution is needed in interpreting them, since answersmight differ depending on who the assessor is and whatmeaning is given to “HSS”. Suggestions for a new orienta-tion of GHIs have started to emerge based on accumulatedcountry-based experience and other complex factors [49].GFATM and GAVI have both included in their fund-

ing programs a stream for HSS, but results have beenmixed: GFATM interrupted its HSS program a year afterits launch, due possibly to a lack of clear definition ofHSS [50]. No clear objective for HSS has been set andeach program has been left to strengthen its own system.GAVI-HSS programs have not taken into account sus-tainability: for example in Burundi, the GAVI-HSS fundswere used inter alia to pay for fuel and salaries fordrivers, in order to transport pregnant women to referralhospitals. Also, initially, an item for HRH capacitystrengthening existed in all GFATM proposals, but thesewere mainly interpreted as in-service training by recipi-ents or used to hire extra staff, without a clear sustain-ability plan. Some of these HRH are now left without asecure position since GHI funding has shrunk.


Similarly, the MEPI and NEPI are projected as exam-ples of PEPFAR platforms for strengthening the sustain-ability of HRH in the countries [22] though the majorityof funding is still clearly earmarked for expanding HIVmanagement and HIV research capacity [31]. The non-coordination between GHIs at global level, impactingnegatively coordination at country level, has alsoemerged as a common feature to all countries, in ourexample of HRH strategies. In an attempt to respond ina coordinated way to health system challenges, a projectto develop a joint HSS funding platform between theWorld Bank, GAVI and GFATM was launched in 2009[51,52]. However, the project has stalled, showing, interalia, the extent to which an agreement on HSS objec-tives and visions is difficult to obtain [53].Our study has some limitations.Qualitative data were collected in a cross-sectional,

manner with different time-periods (Table 1), rising issueof temporal bias while comparing findings.Quantitative data proved difficult or impossible to com-

pare, since HRH categories were not the same acrosscountries, and data were not consistently available orcomplete for the same time periods. The numbers of HRHspecific to HIV were not available; nor were the distribu-tion of HRH according to NGOs/private/public sector.Hence, quantitative data were provided solely for an illus-trative purpose and do not serve as a basis for a rigorousstatistical comparison. Information on non-response ratesto requests for interviews was not available consistentlyacross the 5 countries and was therefore not presented.This could have contributed to an ascertainment bias.Some important institutions such as professional

councils were not interviewed at first, or, even whensolicited, did not reply (Health Professionals Council inSouth Africa). This constitutes a limitation given the rolethat professional councils play in HRH policies.Particular policies adopted to counteract a specific ef-

fect induced by GHIs were difficult to distinguish frommore general policies adopted by the government to ad-dress the overall HRH crisis, which pre-dated GHIs.Finally, given the open nature of the health system, at-

tribution of changes to the influence of GHIs or activ-ists/researchers or countries is an impossible task (andprobably not useful). The truth probably lies somewherebetween these different interpretations, where stream ofinfluences mix with experiences and beliefs and create ageneral direction that actors tend to follow and refinethrough compromise and consensus. This last point re-fers directly to the inherent complexity of the field ofhealth policy and system research, constantly influencedby political and social dynamics [54].Despite these limitations, we believe that this study

provides a unique insight into the complex and slowprocess of policy shift which has occurred in five

countries, between 2007 and 2011, using the example ofHRH policies and their evolution in response to GHIsand related HIV activities. Collectively, the findings pointto some consistent patterns and effects of GHIs on HRHpolicies and the workforce, as well as some context-specific differences, across these five southern Africancountries.

ConclusionThis cross-country paper has shown the differences incountries’ strategies, in response to a common challenge,illustrating the extent to which factors inherent to coun-tries are influential. These findings provide further evi-dence of the importance of country led policies in HSS[55]. Countries are best positioned to assume responsibil-ity on how to best use the global funding available for thepurpose of HSS. HSS proposals should be scrutinizedcarefully according to each country’s context and broaderfactors, since the “one size fits all” strategy has proven itsinefficiency and sometimes its counter-effectiveness in thelong-run. A preliminary impact assessment might be onemechanism, to anticipate any unexpected outcome due toa particular type of funding/policies on the general healthsystem [56]. There is also an urgent need to better definewhat exactly “health system strengthening” means for do-nors, for recipients, but most importantly for health sys-tem users, i.e. for patients.

Competing interestsThe authors declare that they have no competing interests.

Authors’ contributionsAP, TM, EM, RB, JC, IC, TM, BC carried out part of the fieldwork. DS, BC, WVD,RB, GD, UL participated in the study conception and in the design of thestudy. JC drafted the manuscript. JC, AP, WVD, LVL, TM, EM, RB, TM, BC, GD,RB, IC, DS, UL participated to data analysis, interpretation and edited themanuscript. All authors read and approved the final manuscript.

AcknowledgementsThis study was funded by a European Commission project “GHIs in Africa”(grant number INCO-CT-2006-032371).We thank Ms Renata Mares for her useful comments to the paper during itsdesign.

Author details1School of Public Health, Faculty of Community Health Sciences, Universityof the Western Cape, Cape Town, South Africa. 2Unit of International PublicHealth and Biostatistics, Instituto de Higiene e Medicina Tropical, CMDT,WHO Collaborating Centre for Health Workforce Policy and Planning,Universidade Nova de Lisboa, Lisbon, Portugal. 3Eduardo MondlaneUniversity, Maputo, Mozambique. 4Faculty of Health Sciences, NationalUniversity of Lesotho, Maseru, Lesotho. 5Department of Public Health,Institute of Tropical Medicine, Antwerp, Belgium. 6Department ofEpidemiology and Public Health Medicine, Division of Population HealthSciences, Royal College of Surgeons in Ireland, Dublin 2, Ireland.

Received: 24 June 2013 Accepted: 25 September 2013Published: 25 October 2013

References1. Bernstein M, Sessions M: A trickle or a flood: commitments and disbursement for

HIV/AIDS from the global fund, PEPFAR, and the World Bank’ s multi-country AIDSprogram (MAP). Washington DC, USA: Center for Global Development; 2007:1–26.


2. Brugha R: Global health initiatives and public health policy. IntEncyclopedia Public Health 2008, 3:72–81.

3. Komatsu R, Korenromp EL, Low-Beer D, Watt C, Dye C, Steketee RW, Nahlen BL,Lyerla R, Garcia-Calleja JM, Cutler J, Schwartlander B: Lives saved byglobal fund supported HIV/AIDS, tuberculosis and malaria programs:estimation approach and results between 2003 and end-2007.BMC Infect Dis 2010, 10:109.

4. Bendavid E, Bhattacharya J: The President ’s emergency plan for AIDS reliefin Africa : an evaluation of outcomes. Ann Intern Med 2009, 150:688–695.

5. Working together for health. Geneva: The World Health Report of the WorldHealth Organization; 2006.

6. Hirschhorn LR, Oguda L, Fullem A, Dreesch N, Wilson P: Estimating healthworkforce needs for antiretroviral therapy in resource-limited settings.Human Resour Health 2006, 4:1–16.

7. Van Damme W, Kober K, Kegels G: Scaling-up antiretroviral treatment inSouthern African countries with human resource shortage: how willhealth systems adap? Soc Sci Med 2008, 66:2108–2121.

8. Mtonya B, Chizimbi S: Systemwide Effects of the Global Fund in Malawi: FinalReport. Bethesda, USA: The Partners for Health ReformPlus project, AbtAssociates Inc.; 2006:1–54.

9. Biesma RG, Brugha R, Harmer A, Walsh A, Spicer N, Walt G: The effects ofglobal health initiatives on country health systems: a review of theevidence from HIV / AIDS control. Health Policy Plan 2009, 24(4):239–252.

10. Brugha R, Simbaya J, Walsh A, Dicker P, Ndubani P: How HIV / AIDSscale-up has impacted on non- HIV priority services in Zambia.BMC Public Health 2010, 10:540.

11. Hanefeld J, Musheke M: What impact do Global Health Initiatives have onhuman resources for antiretroviral treatment roll-out ? A qualitativepolicy analysis of implementation processes in Zambia. Hum ResourHealth 2009, 7:1–13.

12. Oomman N, Wendt D, Droggitis C: Zeroing. Washington, DC: AIDS Donorsand Africa’s Health Workforce; 2010:1–72.

13. World Health Organization Maximizing Positive Synergies CollaborativeGroup: An assessment of interactions between global health initiativesand country health systems. Lancet 2009, 373:2137–2169.

14. Human Resources for Health: Overcoming the Crisis. Joint Learning Initiative.Boston, USA: Harvard University; 2004.

15. Samb B, Celletti F, Holloway J, Van Damme W, De Cock KM, Dybul M: Rapidexpansion of the health workforce in response to the HIV epidemic.N Engl J Med 2007, 357:2510–2514.

16. Human resources for health strategy 2012 to 2017. South Africa: NationalDepartment of Health; 2012.

17. Politique de développement des ressources humaines. Burundi: Ministère de laSanté; 2010.

18. Scheffler RM, Mahoney CB, Fulton BD, Dal Poz MR, Preker AS: Estimates ofhealth care professional shortages in sub-Saharan Africa by 2015.Health affairs 2009, 28:w849–w862.

19. Wildshut A: Doctors in the public service: too few for too many.HSRC Review 2010, 8(4).

20. Bedelu M, Ford N, Hilderbrand K, Reuter H: Implementing antiretroviraltherapy in rural communities: the Lusikisiki model of decentralized HIV/AIDS care. J Infect Dis 2007, 196(Suppl 3):S464–8.

21. GFATM: The Framework Document. Geneva, Switzerland: The GFATM;2001:1–22.

22. Institute of Medicine: Evaluation of PEPFAR. Washington, DC: The NationalAcademies Press; 2013.

23. Tyrrell AK, Russo G, Dussault G, Ferrinho P: Costing the scaling-up ofhuman resources for health: lessons from Mozambique and GuineaBissau. Human Resour Health 2010, 8:1–10.

24. Burundi: “Die-in” to protest lack of HIV care. http://www.irinnews.org/Report/92363/BURUNDI-Die-in-to-protest-lack-of-HIV-care.

25. Georgeu D, Colvin CJ, Lewin S, Fairall L, Bachmann MO, Uebel K,Zwarenstein M, Draper B, Bateman ED: Implementing nurse-initiated andmanaged antiretroviral treatment (NIMART) in South Africa: a qualitativeprocess evaluation of the STRETCH trial. Implementation Sci 2012, 7:1–13.

26. Fairall L, Bachmann MO, Lombard C, Timmerman V, Uebel K, Zwarenstein M,Boulle A, Georgeu D, Colvin CJ, Lewin S, Faris G, Cornick R, Draper B,Tshabalala M, Kotze E, Van Vuuren C, Steyn D, Chapman R, Bateman E: Taskshifting of antiretroviral treatment from doctors to primary-care nursesin South Africa (STRETCH): a pragmatic, parallel, cluster-randomised trial.Lancet 2012, 380:889–898.

27. Dreyer A: Clinical associate programme: understanding this new level ofHealth Care worker. Johannesburg: Public Health Association of South Africaannual conference; 2011.

28. Introduction of a New Job/work Level of Clinical Associate. South Africa:Natinal Department of Health; 2010.

29. Pillay Y, Baron P: The implementation of PHC re-engineering in South Africa.East London: Public Health Association of South Africa annual conference;2010:1–6.

30. Support for district health plans and their implementation. District HealthSystem News 2011, 3:1–7.

31. Medical Education Partnership Initiative (MEPI). Fogarty International Centerof the national Institutes of Health; 2011.

32. Dohrn J: Nursing Education Partnership Initiative (NEPI ). NYC, USA: ICAP,Columbia School of Public Health; 2011.

33. Sherr K, Mussa A, Chilundo B, Gimbel S, Pfeiffer J, Hagopian A, Gloyd S:Brain drain and health workforce distortions in Mozambique. PloS one2012, 7:e35840.

34. Horton R: Offline: the struggle for leadership. Lancet 2012,380(Dec 8):1977–1978.

35. Launch of the Frontline Healthcare Workers Coalition. http://blogs.state.gov/2012/01/article/launch-frontline-healthcare-workers-coalition.

36. Vujicic M, Weber SE, Nikolic IA, Atun R, Kumar R: An analysis of GAVI, theGlobal Fund and World Bank support for human resources for health indeveloping countries. Health Policy Plann 2012, 27:649–657.

37. Walsh A, Ndubani P, Simbaya J, Dicker P, Brugha R: Task sharing in Zambia:HIV service scale-up compounds the human resource crisis. BMC HealthServ Res 2010, 10:272.

38. Brugha R, Kadzandira J, Simbaya J, Dicker P, Mwapasa V, Walsh A: Healthworkforce responses to global health initiatives funding : a comparisonof Malawi and Zambia. Hum Resour Health 2010, 8:19.

39. Global HIV / AIDS Initiatives and Human Resources for Health in Zambia. GHINpolicy brief; 2009. October.

40. Frenk J, Chen L, Bhutta Z, Cohen J, Crisp N, Evans T, Fineberg H, Garcia P,Ke Y, Kelley P, Kistnasamy B, Meleis A, Naylor D, Pablos-Mendez A, Reddy S,Scrimshaw S, Sepulveda J, Serwadda D, Zurayk H: Health professionals fora new century: transforming education to strengthen health systems inan interdependent world. Lancet 2010, 376:1923–1958.

41. Galichet B, Goeman L, Hill PS, Essengue MS, Hammami N, Porignon D,Kadama P: Linking programmes and systems : lessons from the GAVI healthsystems strengthening window. Trop Med Int Health 2010, 15:208–215.

42. Tangcharoensathien V, Patcharanarumol W: Global health initiatives :opportunities or challenges? Health Policy and Planning 2010, 25:103–105.

43. Chee G, Pielemeier N, Lion A, Connor C: Why differentiating betweenhealth system support and health system strengthening is needed. Int JHealth Plann Mgmt 2013, 28(1):85–94.

44. Duber HC, Coates TJ, Szekeres G, Kaji AH, Lewis RJ: Is there an associationbetween PEPFAR funding and improvement in national health indicatorsin Africa? A retrospective study. J Int AIDS Soc 2010, 13:21.

45. El-sadr WM, De Cock KM: Health systems exist for real people. J AcquirImmune Defic Syndr 2009, 52(1):S1–S2.

46. Palen J, El-sadr W, Phoya A, Imtiaz R, Einterz R, Quain E, Blandford J, Bouey P,Lion A: PEPFAR, health system strengthening, and promoting sustainabilityand country ownership. J Acquir Immune Defic Syndr 2012, 60(Suppl 3):113–119.

47. Car J, Paljärvi T, Car M, Kazeem A, Majeed A, Atun R: Negative health systemeffects of Global Fund’s investments in AIDS, tuberculosis and malaria from2002 to 2009: systematic review. J R Soc Med Sh Rep 2012, 3:1–14.

48. Walensky RP, Kuritzkes DR: The impact of the president’s emergency planfor AIDS relief ( PEPfAR ) beyond HIV and why it remains essential.CID 2010, 50:272–275.

49. Buse K, Tanaka S: Global public-private health partnerships : lessonslearned from ten years of experience and evaluation. Int Dent J 2011,61(Suppl. 2):2–10.

50. Bertolone M: Decision-making in global health initiatives: the example of thediscontinuation of health systems strengthening proposals by the global fund.London, UK: The London School of Economics and Political Science;2007:1–30.

51. England R: The GAVI, Global Fund, and World Bank joint fundingplatform. Lancet 2009, 374:1595–1596.

52. Hill PS, Vermeiren P, Miti K, Ooms G, Van Damme W: The health systemsfunding platform: is this where we thought we were going?Globalization Health 2011, 7:1–16.

http://www.irinnews.org/Report/92363/BURUNDI-Die-in-to-protest-lack-of-HIV-care

http://www.irinnews.org/Report/92363/BURUNDI-Die-in-to-protest-lack-of-HIV-care

http://blogs.state.gov/2012/01/article/launch-frontline-healthcare-workers-coalition

http://blogs.state.gov/2012/01/article/launch-frontline-healthcare-workers-coalition


53. Brown SS, Sen K, Decoster K: The health systems funding platform andworld bank legacy: the gap between rhetoric and reality.Globalization Health 2013, 9:1–7.

54. Gilson L, Hanson K, Sheikh K, Agyepong IA, Ssengooba F, Bennett S:Building the field of health policy and systems research: social sciencematters. PLoS Med 2011, 8:1–6.

55. Borisch B: GHIs and the new dichotomy in health systems. J Public HealthPolicy 2010, 31:100–109.

56. Swanson RC, Bongiovanni A, Bradley E, Murugan V, Sundewall J, Betigeri A,Nyonator F, Cattaneo A, Harless B, Ostrovsky A: Toward a consensus onguiding principles for health systems strengthening. PLoS Med 2010, 7:1–6.

doi:10.1186/1744-8603-9-52Cite this article as: Cailhol et al.: Analysis of human resources for healthstrategies and policies in 5 countries in Sub-Saharan Africa, in responseto GFATM and PEPFAR-funded HIV-activities. Globalization and Health2013 9:52.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

analysis of human resources for health strategies and policies

Documents